-
Notifications
You must be signed in to change notification settings - Fork 5
/
a_note_on_AESOP.txt
51 lines (37 loc) · 2.39 KB
/
a_note_on_AESOP.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
The AESOP (Agents and Emergencies for Simulating Organizations in Python) implementation,
has the capability of reading the schedule from a text file.
The test file is schedule.txt and its contents are read if we have "read_script"
steps activated in the schedule of the model.
If the file is finished, further attempt to "read_schedule" have no effect.
We can read more than one command per step.
If we create a xls file, this is used instead of the txt one.
If schedule.xls exist it is interpreted and translated to schedule.txt,
without the comments.
The script has the capability of broadcasting a method only to a share of the
instances of a class; see in schedule.xls of the project 'basic' (or
in schedule.txt) the row 'all 0.5 dance'
We have a WorldState class, interacting with the agents; at present, in the
'basic' case we have a unique instance of the class, but the code is anyway
built upon a list of any number of instances of the class
The variables managed via WordState have to be added, with their methods, within
the class, following the existing example in project 'basic', where
WorldState has set/get methods for the variable generalMovingProb
We can have more than one category of AESOP agent (behing the scene, currently
a unique Agent.py class exists), each defined in a separate file
(a file can contain also a stand alone agent)
We have also the possibility of creating operating sets, listed in
agOperatingSets.txt, containing the IDs of agents also of different type;
the agents of a set are asked to to something together
We have macro instructions, as separate sheets in schedule.xls, to be invoked as
macro nameOfTheMacro
We can insert empty lines in schedule.xls, to improve readability
the part of agents in a set, effectively acting, can be identified both as an
individual probability - with a real (floating positive number) - or as an
absolute number - expressed as an integer negative number (in case, the acting
agent/s will be extracted randomly from the specified set)
Via spreadsheet, the code has the the capability of reordering a schedule
not ordered upon time, with '# x' rows in macros, with repeated
action over time in a macro; see the example complicated_schedule.xls in school
folder; to use it, rename the file (preserving the existing one) to the default
name: schedule.xls
We have conditional methods, i.e. shakeIf_greenPupil (in scheduleIf.xls)